The ideas of regular data in C++ have a long and instructive history.

Articles in this series

Also see the above in presentation form

Introduction

Consider this type, person that I introduced recently in the series of articles on programming idioms.

1
2
3
4
5
6
7
8
9

struct person
{
  std::string first_name;
  std::string last_name;
  int age{};

  constexpr std::strong_ordering
    operator<=>(const person &) const noexcept = default;
};

I called this a regular data type. Regular is another name for normal, usual. person is type that holds some data related to a person. Behind it’s relative simplicity it’s a very flexible type: it can be copied, moved, compared for equality and order, it will not leak memory.

Regularity is really the benchmark on which other types will be judged i.e. “this other type really can’t be regular because …”

In this series of articles I’m going to cover what I know about the history of the design supporting regular data types. The history is long and contains interesting deliberate and accidental events.

It is a selective history, so that if we look at the types like the person above we can remember: “this is like this because of this reason/choice/accident”.

References

Again, a lot of the ideas in this series of articles come from Elements of Programming by Alexander Stepanov and Paul McJones, from comments done by Alexander Stepanov in recordings published on Youtube and from reading carefully some C++ committee papers.

The “Design and Evolution of C++” by Bjarne Stroustrup is an important source for how ideas that lead to C++ from C came to be. It does stop at about 1994, so it does not cover later developments.